Skip to content

[Model] Support Qwen-VL and Qwen-VL-Chat models with text-only inputs #5710

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
ywang96 merged 2 commits into from
Jun 22, 2024
Merged

[Model] Support Qwen-VL and Qwen-VL-Chat models with text-only inputs #5710

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
5e42e28
[Model] Support Qwen-VL and Qwen-VL-Chat models with text-only inputs
DamonFool Jun 20, 2024
d766681
add TODO
ywang96 Jun 21, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 10 additions & 0 deletions vllm/model_executor/models/qwen.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -28,6 +28,7 @@
from vllm.model_executor.model_loader.weight_utils import default_weight_loader
from vllm.model_executor.sampling_metadata import SamplingMetadata
from vllm.sequence import SamplerOutput
from vllm.utils import print_warning_once


class QWenMLP(nn.Module):
Expand Down Expand Up @@ -288,6 +289,15 @@ def load_weights(self, weights: Iterable[Tuple[str, torch.Tensor]]):
# Skip loading extra bias for GPTQ models.
if name.endswith(".bias") and name not in params_dict:
continue
# Skip loading visual weights to support Qwen-VL models
# in cases with text-only inputs
# TODO: add support for Qwen-VL
if (name not in params_dict
and name.startswith("transformer.visual.")):
print_warning_once(
"Only text inputs are allowed. Images won't be handled "
"until Qwen-VL models are fully supported.")
continue
param = params_dict[name]
weight_loader = getattr(param, "weight_loader",
default_weight_loader)
Expand Down